DK2063111T3 - Process for regulating a wind power plant - Google Patents

Abstract

Method for controlling a wind power plant comprises adjusting the rotational speed of the rotor blade when the limiting value corresponding to the angle of the rotor blade is attained. Preferred Features: The limiting value is a defined rotor blade limiting angle, which is formed by adjusting the rotational speed above the limiting value as a temporary and/or local average value from several rotor blade angles.

Description

The invention relates to a method for controlling a wind power plant according to the preamble of claim 1 and a wind power plant according to the preamble of claim 17.

Known methods are used in wind power plants having a rotor having at least one rotor blade, the rotor blade being arranged at an adjustable rotor blade angle to the rotor. In the known methods, the rotor speed is controlled within a set wind speed range by changing the rotor blade angle to set a fixed power.

For a better understanding, the regulation of a wind power plant is explained below with the aid of the one shown in FIG. 1 shows the power or speed characteristics as a function of the wind speed.

Normal power and rpm characteristics, as they are now predominantly run in wind power plants, have two areas.

The first range of power or rpm characteristic is the lower partial load range UB. This starts at the so-called switching speed vE and ends at the nominal wind speed vN. The switch-on wind speed is the wind speed at which a wind power plant effectively begins to emit power. At wind speeds below the switching speed vE, the power produced by the rotor covers only the required loss power for the drive string and its own consumption. In the lower partial load range, the effect or speed increases with increasing wind speed until the nominal wind speed vN is obtained. Thereby, the rotor blade angle is kept substantially constant and the speed is controlled or controlled via the speed-dependent torque. The power adjusts depending on the wind speed absorbed by the rotor wind speed. In the lower partial load range, the rpm characteristic has two sections a and b, where the section a is formed as a plateau along the lower rpm limit Ωυ. In the second section b, the speed increases linearly depending on the wind speed (operation at the optimum speed speed) until the upper speed limit with the rated speed is reached. The nominal rpm can be obtained at the nominal wind speed, but for the most part the nominal rpm is obtained earlier.

The nominal water velocity vN is the wind speed at which the power of the wind power system corresponds for the first time to the so-called nominal power.

The other area of the power or speed characteristic is the so-called nominal power range, Nv. The nominal power range begins at the nominal wind speed vN and ends at the decoupling wind speed VA, where the rotor speed within the nominal power range is controlled by changing the rotor blade angle to set the rated power. The decoupling wind velocity VA is the highest wind speed at which the plant must be operated at power output. The normal operation of a wind power plant, at which the nominal power is adjusted by changing the rotor blade, is not possible at higher wind speeds than the decoupling wind speed, as in that case there would be damage or even damage to the system due to the additional powerful mechanical load. It is therefore standard instantaneous to decouple wind power plants by obtaining a decoupling wind speed. In general, the power from wind power plants is supplied to the public grid. However, the sudden decoupling of wind power plants at decoupling wind speeds can cause voltage or frequency reductions in the public grid, especially in areas with many power plants. It is therefore sought that the decoupling of the plant does not occur suddenly but by a slow reduction of the power.

In addition, with a reduction of the power, in addition to a compatibility with the mains, it is possible that the plant can be operated on despite an excess of the decoupling wind speed, without damaging the plant due to the increasing wind speed mechanical loads.

A method by which the power and operating speed are controlled after obtaining a decoupling wind speed in dependence on the wind speed is known, for example. from EP 0 847 496 B1. In FIG. 2, for example, a power or rpm characteristic is shown according to this known method, which, in addition to the characteristics discussed above, also has a third region, the upper partial load region OB. This begins with the attainment of a limit wind speed V Limit, which corresponds essentially to the previous decoupling wind speed.

Also known from DE 198 44 258 A1 is a method by which the effect of obtaining a set wind speed is reduced. However, the power reduction is already initiated by obtaining a wind speed below the decoupling wind speed.

It is common for the two methods known in the prior art that the regulation of power takes place in the upper partial load range from the attainment of a set boundary wind speed in dependence on the measured increasing wind speed. In practice, however, the turnover of the regulation of the plant in the upper part load area, depending on measured wind speeds, is considered extremely problematic.

The operation of a wind power plant in the upper part load area means that the wind power system is exposed to high loads due to higher wind speeds. This is partly due to the fact that the air flow is not stationary but turbulent. Turbulence means that the speed at which the wind strikes the rotor surface is not the same at each location of the rotor, but that the wind energy is distributed differently over the entire rotor surface. This means that in one or more areas of the rotor extremely high wind forces can attack. At the same time, in another area or several areas of the rotor, on the other hand, only very small or even no wind forces can be attacked, so that the plant at high turbulence is subjected to very heavy changing loads that can immediately or in the long term damage the plant. The energy fluctuations and thus also the alternating loads are particularly large at high wind speeds.

Therefore, for a power or speed control in the upper partial load range, it is desirable that in addition to the wind power, the changing loads described above in these wind speed ranges are also taken into account due to the turbulence.

Therefore, using the known methods, it is necessary that the measured wind speed which underlies the regulation reflects the actual prevailing wind conditions. However, there has been no practice-oriented method so far with which such wind speed can be reliably determined.

It is known to carry out the wind speed measurement using a Gondola anemometer. However, this is a relatively inaccurate method of measurement, as the measurement is greatly falsified due to the wind rotor. The area of influence of the rotor is at least one rotor diameter in front and 3 rotor diameters behind the rotor. Because of this, an undisturbed measurement is not possible. Second, with this measurement method, it is only possible to assess the measurement conditions for a point-shaped area of the wind field. However, with this known method it is not possible to reproduce the actual turbulent wind speed prevailing over the entire rotor.

From several publications, e.g. from WO 2004/077068 A1 or DE 101 37 272 A1 it is known to use so-called LIDAR or SODAR systems for the three-dimensional measurement of wind conditions in front of the wind power plant. This is described as sensible, especially in the context of wind farms.

So far, these systems have only been used in simplified form since they have not yet been adequately tested. Furthermore, LIDAR and SODAR systems are very expensive, and so far it has not been possible to equip each energy plant with such a system so far that such an application for the actual wind conditions has not been implemented.

Furthermore, from EP 1 230 479 B1 there is known a method in which the mechanical load of the rotor blades is determined by means of sensors placed in the rotor blades, and the particular load values can e.g. is also used to regulate the plant. FIG. 2 in EP 1 230 479 B1 shows a diagram showing the effect as a function of the wind speed, and fig. 3 of EP 1 230 479 B1 shows a diagram showing the measured blade loads as a function of the wind speed.

As shown in FIG. 2 and FIG. 3 of EP 1 230 479 B1 has the blade load in the wind speed range in which according to FIG. 2 of EP 1 230 479 B1 had to start with a reduction in power, a plateau-shaped course. This means that the mechanical loads of the rotor blades due to a change in the rotor blade angle and the associated reduction of the wind attack surface of the rotor blades remain, on average, substantially constant. Because of this, the blade load cannot be utilized as an operating parameter to control the power or rpm characteristic in the upper partial load range, since such control only works with a clear change in the average blade load, and this according to FIG. 3 is not the case. Short-term load peaks can be used to regulate the wind turbine, for example. can ensure that the blade tips of the rotor blades do not collide with the tower. However, adjusting the power or speed by means of load tips is not advantageous, as the power due to compatibility with the mains and the speed due to the high inertia of a wind rotor can only be slowly changed, however, the load points have already been cut off. Because of this, this method has only one option to detect the mechanical load of the rotor blades in the upper partial load range and to take into account the specific data of the control in the upper partial load range. However, a detailed regulation of the plant in the upper partial load range on the basis of the specified load values is not obvious, so that the control can also be carried out only in accordance with the method of EP 1 230 479 in the upper partial load range depending on the wind speed.

A summary of the method described above shows that it has been known so far to regulate the power or speed in the upper partial load range depending on the wind speed. However, since it has not been possible without high production costs to determine a wind speed that reflects the actual wind conditions, the methods described above have the disadvantage that in practice they are only difficult to realize. DE 103 00 733 B3 discloses an operating system for a wind energy system through which a power output for the system is controlled, the wind energy system having a rotor having at least one rotor blade disposed at an adjustable rotor blade angle relative to the rotor and the operating system within a predetermined wind speed range controls the rotor speed while adjusting the rotor blade angle for setting a rated power, and from a defined wind speed dependent limit value reduces the power, the limit value being a defined rotor blade limit angle. US 4,161,658 A discloses a wind energy system operated via a variable pitch angle, serving as an input size to a rotor speed regulator, a torque, a generator speed and a generator power, and a pitch angle and, moreover, a wind speed. The wind speed can be an average wind speed.

The prior art has so far failed to find a method which, in particular for a multimega-megawatt system, enables a power-reduced system operation in a wind speed range between e.g. 25 and 35 m / s or 25 and 40 m / s, without increasing the load of the wind power plant effectively. In particular, at high wind speeds, at which the fluctuations of the wind speed are associated with large fluctuations of the incident energy, the control algorithms known in the state of the art are also very vibrationally susceptible. For this reason, the above-mentioned methods have so far hardly been used in practice.

The reduction in power or rpm should therefore not take place on the basis of the measured wind speed but on the basis of an input parameter which is both physically and control technically easier to judge and partly a better signal for the system load.

The object of the present invention is to provide an improved and more reliable method for regulating a wind power plant after a certain limit value is reached, and for improving compatibility with the mains.

The object is achieved by a method for regulating a wind power plant having the characteristic features of claim 1.

In the proposed method for regulating a wind power plant, the rotor is driven by a substantially constant nominal speed of rotation in a nominal power range following a lower partial load range until a certain limit value is reached. If the limit value is exceeded, the rotor in the upper partial load range is driven by a reduction of the rotor speed to below the nominal power speed.

The term nominal rpm is understood to mean the rpm at which the wind power system reaches its rated power at a nominal wind speed. Against this, a rotor speed is denoted by the nominal power speed, at which the wind power system can deliver a nominal power. The rated power rpm is not necessarily identical to the rated rpm, but may have a value that is in the vicinity of the rated rpm. This range usually includes, but is not limited to, about +/- 10% of the rated speed.

According to the invention, it is envisaged that when the limit value is reached, the rotor speed will be adjusted depending on the angle of the rotor blade.

The present invention has the advantage that, after exceeding a limit value, as the input parameter for the rotation of the rotor speed, as known from the prior art, does not use the measured wind speed whose measurements, as already explained, are subject to considerable difficulties, but the rotor blade angle which can be determined in a simple way. Instead of the rotor speed, the equivalent of the generator according to the invention can be used completely equivalent.

The invention is further based on the recognition that the set rotor blade angle provides significantly better information regarding the load condition of the system than the measured wind speed. This is because the setting of a blade angle is made depending on the rotor power and rotor speed, and therefore not only is the wind speed reflected through a set blade angle, but also other factors are taken into account, such as. current wind direction, air density, vertical and horizontal wind gradients, wind turbulence, machine dynamics, etc ..

The set rotor blade angle therefore reflects the absorbed wind energy integrated throughout the rotor surface, whereas the wind speed measured with an anemometer only transmits the energy supply for a point in the wind field in the area of the anemometer.

Similarly, due to the complex physical conditions against the widely accepted view for a long time, there can also be no clear correlation between a measured wind speed and the rotor blade angle. This has also been confirmed by means of measurements.

In order to clarify the above, reference should be made to FIG. 4. FIG. 4 shows the record of a large number of rotor blade angles in dependence on the wind speeds measured by normal operation of a wind turbine wind turbine system.

As can be seen from this reproduction, a blade angle of e.g. 15 ° both belong to a wind speed of 13 m / s and a wind speed of 26 m / s or also any value between these two wind speeds. In addition to the factors listed above, the large spread in the measured values is also due to the fact that in regulating a wind power plant these are highly dynamic processes, which, however, are of great importance for the current load condition.

FIG. 4 thus confirms that there is in fact no unambiguous relationship between wind speed and blade angle but only a correlation with a rough correlation, which is not sufficient for accurate regulation, which, however, is necessary for power delivery from a large wind power plant in the upper partial load range.

Furthermore, it is clear from FIG. 4 that the blade angle for this reason constitutes a significantly better input parameter than the wind speed, since this is a parameter whose value can be determined in a simple way, whereas the wind speed can have strong oscillations across the rotor surface.

Since the purpose of regulating the plant in the upper part load area is that operation of the plant is also possible at high wind speeds without exposing the plant to danger of overload, it is unconditionally necessary to take into account the above parameters such as wind direction, wind turbulence , etc .. This is made possible by the present invention in a surprisingly simple manner by a blade angle dependent regulation in the upper partial load range.

Some of the aforementioned influences can be safely eliminated by a sufficiently long averaging of the wind speed. However, as will be further illustrated below, a 15 second averaging only reduces the oscillations significantly and several minutes averaging times as would be necessary to establish a halfway acceptable correlation would only render the regulation undesirably sluggish. Impacts, such as The airtightness effect, leads to seasonal fluctuations. Thus, it is e.g. It is known that in places with severe temperature fluctuations between winter and summer, significant air density fluctuations can also occur, which are directly included in the load of the wind energy system. This effect is completely taken into account by a power reduction depending on the blade angle. In addition, the experience-defective adjustment of the relevant control parameter to the altitude-dependent average air density, advantageously avoided, e.g. is needed on mountain stands.

In order to better understand the aforementioned effects, the blade angle in modern wind energy systems is often not only dependent on the rotor power and rotor speed, ie of physical sizes, which in a further sense constitutes the integration of the energy absorbed by the wind over the entire rotor surface, but the blade angle can also be adjusted in dependence on the measured or estimated loads (calculated).

The method of the invention provides that the operation of the rotor blade angle is permanently monitored. Below the limit value, the operating computer regulates the speed so that an optimal energy yield is obtained, ie. that the rotor is operated with the most favorable calculated rpm value for the operating range. Compliance with the calculated speed is usually carried out by a combination of blade setting and generator torque or power control and is sufficiently precisely described in the prior art. After reaching the limit value, a calculated calculation function lowers the calculated value of the rotor speed depending on the rotor blade angle, ie. leaving the nominal power rpm. In connection with the rpm control described so far, it is always a matter of determining a calculated rpm value which, e.g. is used as a control input for the regulation of the blade setting. However, as with normal operation below the limit value, gusts or negative gusts also lead to dynamic oscillations around the calculated rpm, which must then be corrected by a dynamic change of blade angle and possibly by the generator torque.

Furthermore, as is known in the art in operating the system in the nominal power range, the control of the speed is furthermore essentially effected by adjusting the blade angle, ie. that the blade angle is changed depending on the rpm deviation (calculated rpm - actual rpm) to comply with the calculated rpm. In other words, this is a classic control circuit with a simple signal feedback.

In operation in the upper partial load range, according to the present invention, the speed of rotation is also adjusted by adjusting the blade angle. In addition, however, the calculated speed is controlled depending on the blade angle, ie. that not only does the blade angle change depending on the rpm deviation, but each change in blade angle also leads to a change in rpm deviation (by changing the calculated rpm) since the rpm deviation is used to adjust the blade angle, the change in the calculated rpm also causes a subsequent change of rpm . Figuratively speaking, the change in the calculated speed corresponds to a change in the control target. In this case, this is a double signal feedback. Simulations and measurements on prototypes show that this control circuit can be operated stably and that the method according to the invention works without problems when selecting suitable parameters.

Reduction of the rotor speed takes place by means of a calculation function (the speed as a function of the blade angle) eg. in the form of a table or a mathematical function that is determined taking into account the aerodynamic rotor conditions and loads. Therefore, to obtain optimal results, it is necessary that the calculation function is determined iteratively for each blade type and for each plant version either by computer simulation calculations or by field trials, which are admittedly very slow.

A particularly advantageous embodiment of the invention therefore provides that the calculation function of the rpm as a function of the blade angle is determined and / or optimized by simulation calculations taking into account the plant loads in turbulent wind conditions.

A method for determining or optimizing the rpm blade angle characteristic of the control according to the invention for a wind power plant hereby provides that the behavior of the control is investigated, especially with turbulent wind periods containing positive wind speed jumps, ie. that the average wind speed of e.g. 26 m / s suddenly rises to 35 m / s. In these cases, the risk of wind turbine overload is particularly high.

If the resulting loads exceed the permissible loads, then the rpm blade angle characteristic, the control parameter and / or the control algorithm must be changed. Alternatively, of course, it can also be shown that all components of the machine withstand the increased loads. In simulation calculations for a 5 megawatt plant, it has been found hereby that the rotor speed in the range of the limit value must be reduced relatively quickly to keep the loads in the permissible range. In addition, in order not to let the power drop sharply, the torque of the generator must also be increased compared to the rpm-dependent behavior known from the lower partial load operation.

This, above all, more costly method, however, has in the end effect the advantage that the regulation can be carried out so that the occurring loads do not become higher than when the plant is decoupled by reaching the limit value or that they can be better leveled, and therefore, e.g. the plant's reliability is increased and, despite this, a desirable slow decrease in power can be realized due to the compatibility with the mains.

As a limit value, for example, as is known from the prior art, a certain decoupling rate can be used. As the prior art discussion explains in detail, recording the actual prevailing decoupling rate poses a major problem.

According to an advantageous embodiment of the invention, a certain rotor blade limit angle can also be used as a limit value, such as for example. is described in DE 103 00 733 B3.

According to the invention, the rotor blade angle is calculated as a time-weighted average of several rotor blade angles, whereby the selection quality can be greatly influenced by the control quality. The optimum averaging time is dependent on the energy distribution of the turbulent wind and should preferably be between 2-120 s in the usual European locations, especially advantageously between 10 and 60 seconds.

In a multi-megawatt system with a 5 MW power, the use of a moving 15 sec average has been found to be particularly advantageous both in terms of the desired control stability and in compliance with the permissible system loads. To optimize the control algorithm, the sliding means can be approximated in the internal control by means of a PTrled with a time constant of approx. 7.7 sec.

Since the rotor blades in modern wind turbines can be adjusted individually, the blade angle must not only be determined temporally but also locally, ie. that an average value of the individual rotor blade angle is also calculated. In this way, not only arithmetic mean values are predicted for the temporal as well as for the local averaging, but any other mathematical mean formation, e.g. geometric mean values.

The proposed method for blade regulation with measured loads as proposed in the aforementioned EP 1 230 479 as well as a regulation with so-called load estimator leads to particularly advantageous synergy effects in connection with the method according to the invention. The term plant load here includes both loads and deformations.

According to a further advantageous embodiment of the invention, the generator torque of the generator located in the wind turbine can be set above the interface angle depending on the rotor speed, whereby the generator torque according to a further advantageous embodiment of the invention can be reduced with decreasing rotor speed.

Known wind energy systems have a converter which is designed so that the system can be operated at varying speeds. If the rotor speed (or generator speed) is lowered, the converter is usually unable to maintain the generator torque in the nominal torque size. Therefore, depending on the performance of the converter, the generator torque must also be lowered depending on the speed. This can also be achieved by the presentation of a table function or a mathematical function, and as is already normal for the lower partial load operation in the area of the optimum fast run rate.

In this connection, it is obvious as a beginning to apply the context used for the lower part load area, which is frequently maintained in the form of a table function, the so-called "look-up table", in the control.

However, the load characteristic required for load reasons would, in conjunction with the partial load torque characteristic, lead to an undesirable sharp power reduction in the range of the interface angle. Therefore, it is essential for the desired soft and smooth power reduction that the generator torque is set at significantly higher values than is known from the lower partial load range.

Therefore, a particularly advantageous embodiment of the invention foresees that the adjustment of the generator torque is also made depending on the blade angle. In this way, it is particularly simple to add the function determined to the partial load area, e.g. a table or mathematical context, as well as the adjustment above the rotor blade boundary angle, but this is adjusted depending on the rotor blade angle by means of a correction value so that the desired soft and even power reduction is achieved.

The use of a corrected torque has the advantage that a universal regulator is used for both partial load areas. Since the blade angles in the lower partial load range are always close to zero degrees, the blade angle-dependent correction is not introduced here. The following Table 1 shows, for example, the correction of the torque of a 5 MW plant with a rated speed of 1170 rpm and a nominal torque of approx. 44,900 Nm.

Table 1: Torque-dependent correction of the torque of a 5 MW plant

As can be seen from the table, there is no correction of torque until the rotor blade boundary angle, which is at 25 degrees, is obtained. When exceeding the rotor blade boundary angle, e.g. from a rotor blade angle of 30 degrees, a corresponding adjustment is then made by means of torque correction values.

The second table line shows the blade angle dependent calculated speed, which is made in the form of an upper limitation of the calculated speed. Hereby, it has been found to be particularly advantageous to make the speed limiter intervention particularly soft, since at a blade angle of 23 degrees a speed limit is set just above the nominal speed of the plant example. The interpolation points are interpolated, most simply linearly, but preferably with higher order interpolations. Of course, an interpolation can be done at all table values.

It is particularly advantageous to include the table shown above in the control in relative values instead of the absolute values shown, since their properties can then be essentially transferred to systems of a different size. In connection with higher developed control systems, the generator torque is frequently dynamically controlled, whereby the rotor speed is utilized as an input signal, e.g. for a PID controller. The optimum rpm characteristic required for the lower part load range is often realized by limitations of the PID or also the Pi controller. In this case, the control according to the invention also foresees a blade-angle dependent limitation of the regulator to realize the desired behavior of the torque.

Depending on the available setting parameters, according to a further advantageous embodiment of the invention, instead of the generator torque, the power output can also be adjusted depending on the rotor speed, whereby the power output according to a further advantageous embodiment of the invention can be reduced with decreasing rotor speed.

As already mentioned in the discussion of the state of the art, the control algorithms known from the state of the art which regulate in dependence on the wind speed, especially at high wind speeds, are very vibration sensitive. According to an advantageous embodiment of the invention, it is therefore contemplated that, over the rotor blade boundary angle, control parameters are changed compared to the normal operation of the system by a blade angle dependent adjustment so that the control stability is maintained in the upper partial load range.

This adjustment is due to the experience that the requirements of the regulator change at high wind speeds. For example, if considers a guideline-conforming extreme operating wind gust that increases the wind speed by about 9 m / s, then this equally large wind speed change contains a 24% higher change in energy content when it occurs at an average wind of 32 m / s compared to 25 m / s. . These increased energy fluctuations cause e.g. more powerful rpm fluctuations than they occur in normal operation. If one would react to these powerful rpm fluctuations with the same blade angle change as in normal power operation, the machine would be subject to fluctuations due to the over-reaction. Furthermore, it has long been known that the dependence on the change of the rotor speed of the change of blade angle (sensitivity) increases with increasing blade angle. Both effects reinforce each other and lead to the requirement to reduce pitch gain blade angle dependent on the target known from normal operation. Thus, according to a further advantageous embodiment of the invention, it is envisaged to reduce the so-called pitch gain at blade angles above 30 degrees to less than 20% of the nominal gain, ie. the pitch gain under normal wind. The pitch gain is a control parameter that is scaled in dependence on the blade angle change to be implemented by the current speed deviation. With a simple proportional regulator, this is the proportionality factor between control deviation and change of the set value, ie. that in the specific case of speed control using blade angle change, the blade angle change is calculated by multiplying the speed deviation by the pitch gain. (In fact, conditions are a bit more complicated, since it is not a pure P controller but a PID controller, but also the amplifications of the integral and differential ratios are scaled using the pitch gain).

Thus, it is suggested that the pitch gain at the large blade angles constitutes less than one-fifth of the pitch gain under normal winds, as it has been found in simulation calculations that the said control instabilities, such as oscillations, can be eliminated.

For the reasons mentioned above, it is advantageous to reduce or increase other control parameters depending on the blade angle, such as, for example, the pitch and plant. permissible limit values for the permissible rpm range or pitch rates.

Furthermore, the present invention also relates to a wind power plant having a rotor blade disposed at an adjustable rotor blade angle relative to the rotor, whereby the wind power plant according to the invention can be operated by a method according to claims 1 to 15. In the following, the invention will be further explained by means of several depictions. In the illustrations: FIG. 1: a schematic representation of a power and rpm characteristic with a lower partial load range UB and a nominal power range NB; FIG. 2: a schematic representation of a power and rpm characteristic with a lower partial load range UB, a nominal power range NB and a upper partial load range OB; FIG. 3: a schematic representation of a power, torque and rpm characteristic depending on the rotor blade angle; FIG. Fig. 4: Diagram of the rotor blade angles plotted depending on the wind speed; 5: diagram with simulated measurement data for the generator speed according to the rotor blade angle and depending on the wind speed, fig. 6: a rpm blade angle characteristic of a wind power plant; FIG. 7: a converter characteristic; FIG. Figure 8: Diagram: Record of torque / speed depending on rotor blade angle.

FIG. 1 has already been explained in the discussion of the prior art and, therefore, should not be discussed again at this point.

Also fig. 2 is already explained in the prior art, so that at this point only some additional remarks must be made.

FIG. 2, in the outermost partial load region UB, a fully drawn line shows a rpm characteristic which leads to an optimum energy yield at a limited fixed rpm range between Ωυ and ΩΝ. The speed of control is usually carried out by means of a P1 or PID regulator acting on the generator torque, for which the torque for obtaining the operating speed ramp at an optimum speed limit is limited depending on the speed. Alternatively, as mentioned above, for obtaining a particularly simple control algorithm, a simple rpm-torque dependence, e.g. in the form of a table function. Such a fixed speed torque characteristic would, for example, lead to the line speed inserted in dotted line. The horizontal plateau must be replaced by flat ramps, as there is otherwise no clear and inverse context for a table function. The steep ramp, which can be seen at the very beginning of the rpm characteristic, for example, serves to avoid a tower resonance and is in the example shown below the switching speed vE. However, such a steep ramp can also be completely foreseen in normal operation.

Since the effect according to the physical fundamental equations can be determined from the product of the rpm and torque, it is already apparent from fig. 2, that the torque range UB applicable to the lower partial load range UB must be modified in the upper partial load range OB since a higher power is obtained at the same speed. In the discussion of FIG. 8 we will go into this context more closely.

FIG. 3 shows a power and torque and torque characteristic depending on the rotor blade angle υ. As has already been explained in the discussion of the state of the art, the rotor plate angle in the lower partial load range is substantially constant at nearly 0 degrees. Upon obtaining the rated power, the rated power range OB begins and the rotor blade angle is changed so that the system is operated at rated power, rated torque and rated power rpm. Upon obtaining a rotor blade limit angle u Limit, a soft reduction of the rpm Ω, the torque M and hence also the power P begins, whereby the rpm and torque or power are determined according to the invention depending on the increase of the blade angle. The rpm, torque and power are shown relatively and can therefore also be put in this form in a table in the control.

FIG. 4 has already been explained by the description of the advantages of the invention and therefore should not be discussed again at this point.

FIG. Figure 5 shows the record of a large number of registered generator rpm Ω determined by simulation calculations for a 5 MW plant as a function of wind speed. Since the algorithm for determining the speed as described above works with the sliding 15 sec average of the blade angle, the wind speed was also subjected to a sliding 15 sec average value formation to establish a comparability.

The record of the rpm as a function of the wind speed shows in a very obvious way that the rpm and thus also the power is not reduced depending on the wind speed but depending on the rotor blade angle, which, as already shown in fig. 4, does not have a clear correlation with the wind speed.

FIG. 5 shows that in the control method according to the invention, despite a 15 second average value formation, a rpm of e.g. 900 rpm be associated with a wind speed of between 23 m / s and 44 m / s, or conversely, a wind speed of e.g. 33 m / s can correspond to a speed of from 680 to 1070 rpm, and thus the predominant part of the entire range of the speed, for example a 5 megawatt system which is between 650 and 1660 rpm.

Simulation calculations have shown that precisely the release from the state-of-the-art dependence on wind speed and the introduction of the blade angle according to the invention as a regulating input for rpm and power reduction not only improves the control characteristics, but as described only now makes sense. Reasonably, in this context, it is certainly possible to carry out the power reduction in dependence on the wind speed, but which would render the control of the aforementioned regulatory instabilities and increased loads with this approach uneconomical.

FIG. 6 shows an example of a rpm blade angle characteristic of a system with a rated rpm of 1170 rpm. In the present case, there is to start from there an interface angle of about 24 degrees to reduce the speed. At about 50 degrees a speed of 1000 rpm is obtained. In the case shown, the characteristic is linear. Depending on the loads or desired power characteristics available, it is in practice advantageous to set non-linear as well as progressive or degressive characteristics.

FIG. 7 shows an example of the converter characteristic. The permissible generator torque is recorded as a function of the generator speed. If the energy yield for power-reduced operation is to be maximized, the characteristics will follow '' sharply 'as far as possible. If, first and foremost, a reduction of the operating string loads is desired (eg the drive torque), it may be advantageous to significantly underestimate the characteristic, whereby naturally less energy is supplied to the mains. Exceeding the converter characteristic is only possible for a very short time so as not to jeopardize the converter.

FIG. 8 shows the record of the generator speed (the speed) and the generator torque depending on the blade angle using the values shown in Table 1. The run-off curve for the rpm torque dependence used for the lower partial load range is denoted by M_characteristic and the determined calculated value in the control method according to the invention with blade-angle-dependent torque correction is denoted by M_. The torque correction values shown in Table 1 are shown in the form of the deviation between M_characteristic and M_calculation.

Claims (15)

A method for regulating a wind power plant with a rotor having at least one rotor blade disposed at an adjustable rotor blade angle to the rotor operating in a rated power range (NB) following a lower partial load range (UB) until obtaining a certain threshold value is driven by a substantially constant nominal speed of rotation (ΩΝ) and, by exceeding the limit value in an upper partial load range (OB), is driven by a reduction of rotor speed (Ω) below the nominal power speed (ΩΝ), with the rotor speed ( Ω), after reaching the limit value, is set depending on the rotor blade angle, characterized in that the rotor blade angle for adjusting the rotor speed (Ω) above the limit value is calculated as temporal and / or local mean from the multiple rotor blade angle. Method according to claim 1, characterized in that the limit value is a fixed rotor blade limit angle. The method according to claim 1, characterized in that the temporal mean corresponds to a sliding 10 sec to 60 sec average. Method according to the preceding claim, characterized in that the temporal mean corresponds to a sliding 15 sec average. Method according to any one of the preceding claims, characterized in that the rotor blade angle of the wind energy system is controlled depending on measured or estimated plant loads. Method according to any one of the preceding claims, characterized in that the rotor speed (Ω) above the limit value is reduced with increasing rotor blade angle. Method according to any one of the preceding claims, characterized in that the generator torque (M) of the generator located in the wind energy system above the limit value is set depending on the rotor speed (Ω). Method according to the preceding claim, characterized in that the generator torque (M) is reduced by decreasing rotor speed (Ω). Method according to any one of the preceding claims, characterized in that the generator torque (M) is controlled in dependence on the rotor blade angle. Method according to the preceding claim, characterized in that the setting of the generator torque (M) in the upper and possibly also in the lower partial load range takes place depending on the rotor speed (Ω) and the blade angle. Method according to any one of the preceding claims, characterized in that the output of the wind energy system above the limit value is set depending on the rotor speed (Ω). Method according to the preceding claim, characterized in that the power output is reduced with increasing rotor speed (Ω). Method according to any of the preceding claims, characterized in that the control parameters are changed over the limit value compared to the normal operation of the plant by a blade angle dependent adjustment so that the control stability is maintained in the upper partial load range (OB). Method according to the preceding claim, characterized in that a control parameter is the pitch gain. Method according to the preceding claim, characterized in that the pitch gain at blade angles above 30 degrees represents less than 20% of the nominal gain, and / or at blade angles above 40 degrees constitute less than 16% of the nominal gain.